智能故障诊断研究与发展

在综合大量文献基础上,对国内外智能故障诊断的研究现状进行了述评和归纳,同时指出了目前各种智能诊断方法和技术的特点及局限性.这些方法和技术包括:基于规则的智能诊断、基于模型知识的智能诊断、基于神经网络的智能诊断、基于案例的智能诊断、基于行为的智能诊断.对基于神经网络的智能诊断,具体讨论了模式识别的智能故障诊断神经网络、故障预测的神经网络、基于神经网络的智能诊断专家系统.最后指出了智能故障诊断今后的发展趋势,即集成型智能故障诊断系统、基于机器学习的智能故障诊断系统、由单机诊断到远程分布式全系统智能诊断.

Research and Development of Intelligent System for Fault Diagnosis

Based on a lot of documents and references, the actual state of the research on intelligent diagnosis system is reviewed in and out of abroad, the advantages and limitation of all sorts of intelligent fault diagnosis methods are analyzed. The methods include intelligent diagnosis methods based on rule, knowledge model, nerval network, case and action. Aiming at intelligent diagnosis based on nerval network, the intelligent fault diagnosis with nerval network in model identification, fault forecast of nerval network and the expert system for the intelligent fault diagnosis with nerval network. In the end, the trend for the development of the intelligent fault diagnosis, the integrated intelligent fault diagnosis system is presented. The intelligent fault diagnosis from single diagnosis to remote diagnosis of fully distributed system is introduced.